1. Michael Chui, Markus Löffler, and Roger Roberts
Michael Chui is a
senior fellow with
the McKinsey Global
Institute, Markus
Löffler is a principal
in McKinsey’s
Stuttgart office,
and Roger Roberts
is a principal in the
Silicon Valley office.
1
In most organizations, information travels along familiar routes.
Proprietary information is lodged in databases and analyzed in
reports and then rises up the management chain. Information also
originates externally—gathered from public sources, harvested from
the Internet, or purchased from information suppliers.
But the predictable pathways of information are changing: the
physical world itself is becoming a type of information system.
In what’s called the Internet of Things, sensors and tiny devices
(actuators) embedded in physical objects—from roadways to
pacemakers—are linked through wired and wireless networks, often
using the same Internet Protocol (IP) that connects the Internet.
These networks churn out huge volumes of data that flow to computers
for analysis. When objects can both sense the environment and
communicate, they become tools for understanding complexity and
responding to it swiftly. What’s revolutionary in all this is that these
physical information systems are now beginning to be deployed, and
some of them even work largely without human intervention.
Pill-shaped microcameras already traverse the human digestive tract
and send back thousands of images to pinpoint sources of illness.
Precision farming equipment with wireless links to data collected
from remote satellites and ground sensors can take into account
crop conditions and adjust the way each individual part of a field is
More objects are becoming embedded with sensors
and gaining the ability to communicate. The resulting
information networks promise to create new business
models, improve business processes, and reduce
costs and risks.
The Internet of Things

2. McKinsey Quarterly 2010 Number 22
farmed—for instance, by spreading extra fertilizer on areas that need
more nutrients. Billboards in Japan peer back at passersby, assessing
how they fit consumer profiles, and instantly change displayed
messages based on those assessments.
Yes, there are traces of futurism in some of this and early warnings
for companies too. Business models based on today’s largely static
information architectures face challenges as new ways of creating
value arise. When a customer’s buying preferences are sensed in real
time at a specific location, dynamic pricing may increase the odds
of a purchase. Knowing how often or intensively a product is used
can create additional options—usage fees rather than outright sale,
for example. Manufacturing processes studded with a multitude of
sensors can be controlled more precisely, raising efficiency. And when
operating environments are monitored continuously for hazards or
when objects can take corrective action to avoid damage, risks and
costs diminish. Companies that take advantage of these capabilities
stand to gain against competitors that don’t.
The widespread adoption of the Internet of Things will take time, but
the time line is advancing thanks to improvements in underlying
technologies. Advances in wireless networking technology and the
greater standardization of communications protocols make it possible
to collect data from these sensors almost anywhere at any time. Ever-
smaller silicon chips for this purpose are gaining new capabilities,
while costs, following the pattern of Moore’s Law, are falling. Massive
increases in storage and computing power, some of it available via
cloud computing, make number crunching possible at very large scale
and at declining cost.
Tracking behavior
Monitoring the behavior of persons,
things, or data through space and time.
Examples:
Presence-based advertising
and payments based on locations of
consumers
Inventory and supply chain monitoring
and management
1
Information and analysis
Putting the Internet of Things to work

3. 3The Internet of Things
None of this is news to technology companies and those on the
frontier of adoption. But as these technologies mature, the range of
corporate deployments will increase. Now is the time for executives
across all industries to structure their thoughts about the potential
impact and opportunities likely to develop from the Internet of
Things. We see six distinct types of emerging applications, which fall
in two broad categories: first, information and analysis and, second,
automation and control.
Information and analysis
As the new networks link data from products, company assets, or the
operating environment, they will generate better information and
analysis, which can enhance decision making significantly. Some
organizations are starting to deploy these applications in targeted
areas, while more radical and demanding uses are still in the
conceptual or experimental stages.
1. Tracking behavior
When products are embedded with sensors, companies can track the
movements of these products and even monitor interactions with
them. Business models can be fine-tuned to take advantage of this
behavioral data. Some insurance companies, for example, are offering
to install location sensors in customers’ cars. That allows these
companies to base the price of policies on how a car is driven as well
as where it travels. Pricing can be customized to the actual risks of
operating a vehicle rather than based on proxies such as a driver’s age,
gender, or place of residence.
Or consider the possibilities when sensors and network connections
are embedded in a rental car: it can be leased for short time spans
Enhanced situational
awareness
Achieving real-time awareness of
physical environment.
Example:
Sniper detection using direction of
sound to locate shooters
2

4. McKinsey Quarterly 2010 Number 24
to registered members of a car service, rental centers become
unnecessary, and each car’s use can be optimized for higher revenues.
Zipcar has pioneered this model, and more established car rental
companies are starting to follow. In retailing, sensors that note
shoppers’ profile data (stored in their membership cards) can help
close purchases by providing additional information or offering
discounts at the point of sale. Market leaders such as Tesco are at the
forefront of these uses.
In the business-to-business marketplace, one well-known application
of the Internet of Things involves using sensors to track RFID (radio-
frequency identification) tags placed on products moving through
supply chains, thus improving inventory management while reducing
working capital and logistics costs. The range of possible uses for
tracking is expanding. In the aviation industry, sensor technologies
are spurring new business models. Manufacturers of jet engines retain
ownership of their products while charging airlines for the amount
of thrust used. Airplane manufacturers are building airframes with
networked sensors that send continuous data on product wear and
tear to their computers, allowing for proactive maintenance and
reducing unplanned downtime.
2. Enhanced situational awareness
Data from large numbers of sensors, deployed in infrastructure (such
as roads and buildings) or to report on environmental conditions
(including soil moisture, ocean currents, or weather), can give decision
makers a heightened awareness of real-time events, particularly
when the sensors are used with advanced display or visualization
technologies.
Security personnel, for instance, can use sensor networks that
combine video, audio, and vibration detectors to spot unauthorized
Sensor-driven decision
analytics
Assisting human decision making through
deep analysis and data visualization
Examples:
Oil field site planning with 3D visualization
and simulation
Continuous monitoring of chronic
diseases to help doctors determine best
treatments
3

5. 5The Internet of Things
individuals who enter restricted areas. Some advanced security
systems already use elements of these technologies, but more far-
reaching applications are in the works as sensors become smaller and
more powerful, and software systems more adept at analyzing and
displaying captured information. Logistics managers for airlines and
trucking lines already are tapping some early capabilities to get up-
to-the-second knowledge of weather conditions, traffic patterns, and
vehicle locations. In this way, these managers are increasing their
ability to make constant routing adjustments that reduce congestion
costs and increase a network’s effective capacity. In another
application, law-enforcement officers can get instantaneous data from
sonic sensors that are able to pinpoint the location of gunfire.
3. Sensor-driven decision analytics
The Internet of Things also can support longer-range, more complex
human planning and decision making. The technology requirements—
tremendous storage and computing resources linked with advanced
software systems that generate a variety of graphical displays for
analyzing data—rise accordingly.
In the oil and gas industry, for instance, the next phase of exploration
and development could rely on extensive sensor networks placed in
the earth’s crust to produce more accurate readings of the location,
structure, and dimensions of potential fields than current data-driven
methods allow. The payoff: lower development costs and improved oil
flows.
As for retailing, some companies are studying ways to gather and
process data from thousands of shoppers as they journey through
stores. Sensor readings and videos note how long they linger at
individual displays and record what they ultimately buy. Simulations
based on this data will help to increase revenues by optimizing retail
layouts.
In health care, sensors and data links offer possibilities for monitoring
a patient’s behavior and symptoms in real time and at relatively low
cost, allowing physicians to better diagnose disease and prescribe
tailored treatment regimens. Patients with chronic illnesses, for
example, have been outfitted with sensors in a small number of
health care trials currently under way, so that their conditions can be
monitored continuously as they go about their daily activities. One
such trial has enrolled patients with congestive heart failure. These
patients are typically monitored only during periodic physician office
visits for weight, blood pressure, and heart rate and rhythm. Sensors
placed on the patient can now monitor many of these signs remotely
and continuously, giving practitioners early warning of conditions that
would otherwise lead to unplanned hospitalizations and expensive

6. McKinsey Quarterly 2010 Number 26
Automation and control
emergency care. Better management of congestive heart failure alone
could reduce hospitalization and treatment costs by a billion dollars
annually in the United States.
Automation and control
Making data the basis for automation and control means converting
the data and analysis collected through the Internet of Things
into instructions that feed back through the network to actuators
that in turn modify processes. Closing the loop from data to
automated applications can raise productivity, as systems that
adjust automatically to complex situations make many human
interventions unnecessary. Early adopters are ushering in relatively
basic applications that provide a fairly immediate payoff. Advanced
automated systems will be adopted by organizations as these
technologies develop further.
1. Process optimization
The Internet of Things is opening new frontiers for improving
processes. Some industries, such as chemical production, are
installing legions of sensors to bring much greater granularity to
monitoring. These sensors feed data to computers, which in turn
analyze them and then send signals to actuators that adjust processes—
for example, by modifying ingredient mixtures, temperatures, or
pressures. Sensors and actuators can also be used to change the
position of a physical object as it moves down an assembly line,
ensuring that it arrives at machine tools in an optimum position
(small deviations in the position of work in process can jam or even
damage machine tools). This improved instrumentation, multiplied
hundreds of times during an entire process, allows for major
reductions in waste, energy costs, and human intervention.
Process optimization
Automated control of closed (self-
contained) systems
Example:
Maximization of lime kiln throughput via
wireless sensors
Continuous, precise adjustments in
manufacturing lines
1
Putting the Internet of Things to work

7. 7The Internet of Things
In the pulp and paper industry, for example, the need for frequent
manual temperature adjustments in lime kilns limits productivity
gains. One company raised production 5 percent by using embedded
temperature sensors whose data is used to automatically adjust a kiln
flame’s shape and intensity. Reducing temperature variance to near
zero improved product quality and eliminated the need for frequent
operator intervention.
2. Optimized resource consumption
Networked sensors and automated feedback mechanisms can change
usage patterns for scarce resources, including energy and water, often
by enabling more dynamic pricing. Utilities such as Enel in Italy and
Pacific Gas and Electric (PG&E) in the United States, for example,
are deploying “smart” meters that provide residential and industrial
customers with visual displays showing energy usage and the real-
time costs of providing it. (The traditional residential fixed-price-
per-kilowatt-hour billing masks the fact that the cost of producing
energy varies substantially throughout the day.) Based on time-of-use
pricing and better information residential consumers could shut down
air conditioners or delay running dishwashers during peak times.
Commercial customers can shift energy-intensive processes and
production away from high-priced periods of peak energy demand to
low-priced off-peak hours.
Data centers, which are among the fastest-growing segments of global
energy demand, are starting to adopt power-management techniques
tied to information feedback. Power consumption is often half of a
typical facility’s total lifetime cost, but most managers lack a detailed
view of energy consumption patterns. Getting such a view isn’t easy,
since the energy usage of servers spikes at various times, depending
on workloads. Furthermore, many servers draw some power 24/7 but
Optimized resource
consumption
Control of consumption to optimize
resource use across network
Example:
Smart meters and energy grids that
match loads and generation capacity in
order to lower costs
Data-center management to optimize
energy, storage, and processor
utilization
2

8. McKinsey Quarterly 2010 Number 28
are used mostly at minimal capacity, since they are tied to specific
operations. Manufacturers have developed sensors that monitor each
server’s power use, employing software that balances computing loads
and eliminates the need for underused servers and storage devices.
Greenfield data centers already are adopting such technologies, which
could become standard features of data center infrastructure within a
few years.
3. Complex autonomous systems
The most demanding use of the Internet of Things involves the rapid,
real-time sensing of unpredictable conditions and instantaneous
responses guided by automated systems. This kind of machine
decision making mimics human reactions, though at vastly enhanced
performance levels. The automobile industry, for instance, is
stepping up the development of systems that can detect imminent
collisions and take evasive action. Certain basic applications, such
as automatic braking systems, are available in high-end autos. The
potential accident reduction savings flowing from wider deployment
could surpass $100 billion annually. Some companies and research
organizations are experimenting with a form of automotive autopilot
for networked vehicles driven in coordinated patterns at highway
speeds. This technology would reduce the number of “phantom jams”
caused by small disturbances (such as suddenly illuminated brake
lights) that cascade into traffic bottlenecks.
Scientists in other industries are testing swarms of robots that
maintain facilities or clean up toxic waste, and systems under study
in the defense sector would coordinate the movements of groups
of unmanned aircraft. While such autonomous systems will be
challenging to develop and perfect, they promise major gains in safety,
risk, and costs. These experiments could also spur fresh thinking
Complex autonomous
systems
Automated control in open environments
with great uncertainty
Example:
Collision avoidance systems to sense
objects and automatically apply brake
Clean up of hazardous materials through
the use of swarms of robots
3

9. 9The Internet of Things
about how to tackle tasks in inhospitable physical environments (such
as deep water, wars, and contaminated areas) that are difficult or
dangerous for humans.
What comes next?
The Internet of Things has great promise, yet business, policy, and
technical challenges must be tackled before these systems are widely
embraced. Early adopters will need to prove that the new sensor-
driven business models create
superior value. Industry groups
and government regulators
should study rules on data
privacy and data security,
particularly for uses that touch on
sensitive consumer information.
Legal liability frameworks for
the bad decisions of automated systems will have to be established by
governments, companies, and risk analysts, in consort with insurers.
On the technology side, the cost of sensors and actuators must fall
to levels that will spark widespread use. Networking technologies
and the standards that support them must evolve to the point where
data can flow freely among sensors, computers, and actuators.
Software to aggregate and analyze data, as well as graphic display
techniques, must improve to the point where huge volumes of data
can be absorbed by human decision makers or synthesized to guide
automated systems more appropriately.
Within companies, big changes in information patterns will have
implications for organizational structures, as well as for the way
decisions are made, operations are managed, and processes are
conceived. Product development, for example, will need to reflect far
greater possibilities for capturing and analyzing information.
Companies can begin taking steps now to position themselves for
these changes by using the new technologies to optimize business
processes in which traditional approaches have not brought
satisfactory returns. Energy consumption efficiency and process
optimization are good early targets. Experiments with the emerging
technologies should be conducted in development labs and in small-
scale pilot trials, and established companies can seek partnerships
with innovative technology suppliers creating Internet-of-Things
capabilities for target industries.
The authors wish
to thank their
McKinsey colleagues
Naveen Sastry,
James Manyika, and
Jacques Bughin for
their substantial
contributions to this
article.
Related articles on mckinseyquarterly.com
Eight technology trends to watch
How to make Web 2.0 work